Inside the "capacity auction" of the largest power grid in the United States: Without price controls, the AI boom would have caused electricity prices to rise by another 60%

The latest capacity auction results from PJM Interconnection, the largest grid operator in the United States, reveal a harsh reality: the surge in artificial intelligence and data centers is pushing the U.S. power system to its limits.

In the 2027/2028 base residual auction that ended on December 17, the price of generation capacity in the PJM market soared to $333.40 per megawatt-day (MW-day). This price not only broke the record set in July but also reached the price cap approved by the Federal Energy Regulatory Commission (FERC). In contrast, the average price from 2017 to 2024 was only $108 per MW-day, highlighting that the competition for power resources has entered a heated stage.

However, the most shocking data in this auction report is hidden in the appendix. The report details that if the price cap were removed, the simulated market clearing price would actually reach $529.80 per MW-day. This means that in a completely unregulated market environment, the enormous demand brought by data centers would push electricity prices up nearly 60% above the current "cap price."

Goldman Sachs noted in its analysis of this auction that while the prices met expectations, the amount of supply procured once again failed to meet reliability requirements, exacerbating concerns about resource adequacy in the market. With declining reserve margins and a lack of new capacity, the U.S. electricity market is facing a difficult trade-off between "AI development" and "grid stability," with significant upward risks to future electricity bills.

Surge in Data Center Demand and Supply Bottlenecks

The expansion of the U.S. data center industry is significantly tightening the electricity market. Data shows that in November, the power demand capacity of U.S. data centers increased by 1.6 GW (a month-on-month increase of 4%), far exceeding the average monthly growth of 0.26 GW from 2017 to 2024, bringing the total capacity of U.S. data centers to 44.6 GW.

The PJM market covers key areas, including Virginia's "Data Center Alley," and is at the eye of this demand surge storm. Among all regional electricity markets, PJM had the largest new capacity addition, reaching 0.45 GW, with Virginia alone contributing 0.27 GW. This rapid growth in demand directly led to soaring real-time electricity prices last summer and drove up capacity prices.

Despite strong demand, the supply response has been unusually slow. Goldman noted that while Trump promised to revive nuclear energy, the actual new capacity coming online has been slow. In this auction, only about 350.7 MW of new generation and about 423.6 MW of upgraded capacity were procured, with a total cleared capacity of approximately 134,747 MW. This reflects that even with high prices, there has not been an effective stimulus for a large amount of new capacity to enter the auction process

The Hidden True Costs

The core risk revealed by this PJM auction lies in the "shadow price." Simulation data in the report's appendix shows that without regulatory price caps and floors, all prices for 2027/2028 (except for the DOM LDA region) would clear at $529.80 per megawatt-day.

This figure is not only 60% higher than the actual clearing price but also significantly above the $388.57 per megawatt-day from the 2026/2027 simulated auction. This means that under the same conditions, the expansion of data centers has potentially increased electricity costs by 40% within a year.

Even more concerning is that even under the assumption of being completely unregulated, with prices soaring to $529.80, the additional clearing capacity that the grid could attract is only about 800 MW more than the current level. This indicates that the largest grid system in the United States faces physical constraints—no matter how high the price, it cannot "generate" the tens or even hundreds of gigawatts of additional power needed to meet the cycles of AI and data centers in the short term. It is estimated that if prices are not regulated, the related wholesale electricity costs will increase by $9.9 billion.

Deterioration of Grid Reliability Indicators

The auction results further confirm the deterioration of grid reliability indicators. In this auction, the reserve margin dropped from the previous 18.9% to 14.8%, about 5 percentage points lower than the target reserve margin of 20%. The decline of this key indicator directly signals supply tightness.

Goldman Sachs analysis states that while the increase in effective load-carrying capability (ELCC) from 69% to 92% drove a demand response of 1,847 MW, factors such as the increase in battery resources did not fundamentally reverse the supply shortage situation. In terms of type distribution, the cleared generation resources were basically consistent with the last auction, showing structural stagnation.

Goldman Sachs warns that the next auction scheduled for June 2026 (for the 2028/2029 delivery year) faces significant uncertainty, as this auction has not yet set the FERC-mandated price caps and floors.

If the current simulated uncapped price growth rate is extrapolated, the next PJM auction price could be in the low to mid-range of $600 per megawatt-day, effectively doubling from current levels. Goldman Sachs believes that if the next auction has no cap and the market remains tight, it will have an extremely negative impact on utility bills. As market observers have noted, unless there is a structural change before next year's auction, the U.S. will face a dilemma in 2026 between "AI development" and "air conditioning (AC) electricity."

Risk Warning and Disclaimer

The market has risks, and investment should be cautious. This article does not constitute personal investment advice and does not take into account individual users' specific investment objectives, financial situations, or needs. Users should consider whether any opinions, views, or conclusions in this article are suitable for their specific circumstances. Investing based on this is at one's own risk